On Leveraging Unlabeled Data for Concurrent Positive-Unlabeled Classification and Robust Generation
arXiv:2006.07841
Abstract
The scarcity of class-labeled data is a ubiquitous bottleneck in many machine learning problems. While abundant unlabeled data typically exist and provide a potential solution, it is highly challenging to exploit them. In this paper, we address this problem by leveraging Positive-Unlabeled~(PU) classification and the conditional generation with extra unlabeled data \emph{simultaneously}. We present a novel training framework to jointly target both PU classification and conditional generation when exposed to extra data, especially out-of-distribution unlabeled data, by exploring the interplay between them: 1) enhancing the performance of PU classifiers with the assistance of a novel Classifier-Noise-Invariant Conditional GAN~(CNI-CGAN) that is robust to noisy labels, 2) leveraging extra data with predicted labels from a PU classifier to help the generation. Theoretically, we prove the optimal condition of CNI-CGAN and experimentally, we conducted extensive evaluations on diverse datasets.
Published in International Conference on Image and Graphics (ICIG), 2025
References in corpus (9)
- Conditional Generative Adversarial Nets
- Mutual Mean-Teaching: Pseudo Label Refinery for Unsupervised Domain Adaptation on Person Re-identification
- A Review on Generative Adversarial Networks: Algorithms, Theory, and Applications
- On Leveraging Pretrained GANs for Generation with Limited Data
- Robust Conditional Generative Adversarial Networks
- Countering Noisy Labels By Learning From Auxiliary Clean Labels
- Effective Data Augmentation with Multi-Domain Learning GANs
- Robust conditional GANs under missing or uncertain labels
- Patch-level Neighborhood Interpolation: A General and Effective Graph-based Regularization Strategy